Remanufacturing engines don’t pass the test without running-in process. And most of them never so much as reaching 85% of the standard index. After a careful analysis, it is believed that mainly be caused by being lake of enough running-in process. Then a cycle running-in criterion with low data is designed to verify the speculation, and it even don’t reach 50% of full load level. The results show that remanufacturing engines show a good performance, and they can also reach qualified index. Then, based on the variation of dynamic performance of remanufacturing engines, more engine tests have been done to optimize the running-in criterion. After that, a new running-in criterion established which improves efficiency. Moreover, it is proved effective, too.